Reaction Kinetics for a Novel Flue Gas Cleaning Technology

environmentally intrusive and cost-effective flue gas cleaning technologies need to be developed (1-5). An important goal for new SO2 removal technolo...
0 downloads 0 Views 74KB Size
Environ. Sci. Technol. 2003, 37, 1404-1407

Reaction Kinetics for a Novel Flue Gas Cleaning Technology M A O H O N G F A N , * ,† R O B E R T C . B R O W N , † YAHUI ZHUANG,‡ ADRIENNE T. COOPER,§ AND MASAKATSU NOMURA# Center for Sustainable Environmental Technologies, Iowa State University, Ames, Iowa 50011, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, 100085, P. R. China, Department of Civil and Environmental Engineering, University of South Carolina, Columbia, South Carolina 29208, and Department of Applied Chemistry, Osaka University, Suita, Osaka, 565-0871, Japan

This paper studies the kinetics of the reaction between NaClO3, FeSO4, and NaHSO3, which can potentially be used as an alternative to the conventional lime-limestone process for flue gas desulfurization. The key for the establishment of a kinetic model of the reaction is to find a way to determine concentrations of reactants or products during the reaction. The generation rate of Clduring the reaction was monitored using a Dionex Series 4000i ion chromatograph. Based on the changes of Clconcentrations at the designed initial reaction conditions, reaction orders for each reactant were derived. The reaction orders were determined to be 1.1 for NaClO3, 1.1 for FeSO4, and 1.4 for NaHSO3. The global rate coefficients of the reaction at temperatures ranging from 40 to 80 °C were determined. Furthermore, the preexponential factor and the activation energy in the empirical Arrhenius form of the reaction were derived from the relationship between temperature and its corresponding observed global rate coefficient.

Introduction The harmful effects on the environment of SO2 produced from fossil fuel consumption are becoming a problem worldwide. For a long time, treatment with lime or limestone as an absorbent was the major technique for SO2 removal from flue gas. The disadvantages of this conventional technique include the discharge of a large amount of waste slurry containing calcium sulfate, which either needs to be stored or further treated, along with the generation of large amounts of the greenhouse gas carbon dioxide (CO2) into the air. Furthermore, this technique has led to huge financial burdens for many coal-based power plants. Therefore, less environmentally intrusive and cost-effective flue gas cleaning technologies need to be developed (1-5). An important goal for new SO2 removal technologies is the production of valueadded products. We have conducted research on the absorption of SO2, a coal combustion byproduct, with a solution of FeSO4, an * Corresponding author phone: (515)294-3951; fax: (515)294-3091; e-mail: [email protected]. † Iowa State University. ‡ Chinese Academy of Sciences. § University of South Carolina. # Osaka University. 1404

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 7, 2003

inexpensive and readily available byproduct of several industrial processes (3-5). This SO2 removal method has several advantages: it has high SO2 removal efficiency; it occurs at low temperature; it does not add harmful byproducts to the environment; and it produces polymeric ferric sulfate (PFS), an attractive coagulant for water and wastewater treatment (6, 7). The production of PFS from SO2 with FeSO4 as an adsorption agent basically consists of three steps. The first step is to oxidize Fe2+ into Fe3+ under an acidic environment provided by bubbling SO2 into the reaction system. Another step is to hydrolyze Fe3+ into [Fex(OH)y](3x-y)+ species. Finally, [Fex(OH)y](3x-y)+ and SO42- will combine and form stable, polymeric ferric sulfate characterized by a high basicity. The toxicity of PFS synthesized from SO2 in a simulated flue gas has been evaluated (7). The experimental results have shown that water treated with PFS produced from simulated flue gas had no significant effect on body weight gain or the food and water intakes of tested rats and that no significant difference between the clinical chemistry profile of the PFS-treated rat group and that of control group was found (7). Actually, the SO2 cleaning system studied in this research will be built downstream of the electrostatic precipitator (ESP), where most flue gas desulfurization (FGD) systems are presently located. In most coal-fired plants, ESPs have a particulate removal efficiency greater than 99.5% (8). Therefore, to affect the quality of polymeric ferric sulfate synthesized from flue gas, almost all fly ash containing trace contaminants is captured before reaching the reaction vessels. Also, most sulfate salts of trace contaminants are insoluble or only very slightly soluble in liquid PFS and can be filtered from PFS product. Furthermore, the PFS synthesized from flue gas is used primarily for wastewater treatment, in which the standards for coagulants are less stringent than those for treating drinking water. Therefore, PFS synthesized from flue gas has good prospects for use in wastewater treatment. Despite the attractiveness of the integration of air pollution control with water treatment in this SO2 removal technique (3-5), more research needs to be done before it is developed into a commercial process. For instance, the kinetics of the process, an essential element in optimizing reaction conditions, need to be investigated. The central reactions to this new sulfur removal technology are oxidation of HSO3- and Fe2+ by ClO3-. The oxidation process leads to production of Fe2(SO4)3, which is followed by hydrolysis of Fe3+ to produce PFS. This paper investigates reaction kinetics between NaClO3, FeSO4, and NaHSO3. It is noted that the dissolved O2 can potentially oxidize Fe2+ and SO2 during PFS production. However, it is wellknown that NaClO3 is much stronger than O2 under acidic reaction conditions. Therefore, the effects of dissolved O2 on the oxidation of Fe2+ and SO2 as well as the resulting PFS production can be neglected.

Methods Theory. SO32- and HSO3- may simultaneously exist due to the presence of SO2 in the reaction system. However, the oxidation of Fe2+ requires the presence of H+. Since SO32- is unrelated to the oxidation of Fe2+, only HSO3- can meet this requirement. Consequently, only the reaction between NaClO3, FeSO4, and NaHSO3 needs to be considered for the production of PFS when the pH of the reaction system is less than 7, which can be realized by continuously bubbling SO2 into the reactor. 10.1021/es025828v CCC: $25.00

 2003 American Chemical Society Published on Web 02/15/2003

The reaction between NaClO3, FeSO4, and NaHSO3 can be expressed as follows:

NaClO3 + 2FeSO4 + 2NaHSO3 f NaCl + Na2SO4 + Fe2(SO4)3 + H2O (1) Assuming that cNaClO3,0 cFeSO4,0, and cNaHSO3,0 represent the initial mole concentrations of reactants NaClO3, FeSO4, and NaHSO3, respectively, initial concentration of NaCl is 0, and cCl- is the concentration of NaCl or Cl- generated during the reaction or the quantity of NaClO3 or ClO3- consumed, then the concentrations of all the reactants at any time, t, should be cNaClO3,0 - cCl-, cFeSO4,0 - 2cCl-, and cNaHSO3,0 - 2cCl-, respectively, and the increasing rate of cCl-can be expressed as follows

dcCl) k(cNaClO3,0 - cCl-)n1(cFeSO4,0 dt 2cCl-)n2(cNaHSO3,0 - 2cCl-)n3 (2) where k represents the global rate coefficient and exponents n1, n2, and n3 are the orders with respect to reactants NaClO3, FeSO4, and NaHSO3. Reaction orders n1, n2, and n3 are not necessarily integers because reaction 1 is not an elementary reaction. Special initial reactant concentrations need to be designed in order to derive reaction orders n1, n2, and n3. For example, when cNaClO3,0 and cFeSO4,0 are much larger than cNaHSO3,0, cNaClO3, and cFeSO4, at any time t of the reaction 1 can be considered as constants, or

cNaClO3, ) cNaClO3,0 - cCl- ≈ cNaClO3,0

(3)

cFeSO4 ) cFeSO4,0 - 2cCl- ≈ cFeSO4,0

(4)

and then eq 2 can be simplified as

dcCl) k3(cNaHSO3,0 - 2cCl-)n3 dt

(5)

n2 n1 k3 ) kC NaClO ‚C FeSO 3,0 4,0

(6)

where

dcCl) n3 log(cNaHSO3,0 - 2cCl-) + log k3 dt

dcCl) k(cNaClO3,0 - cCl-)n1‚2n2‚(cNaClO3,0 dt cCl-)n2‚2n3(cNaClO3,0 - cCl-)n3 ) 2n2+n3‚k‚(cNaClO3.0 - cCl-)n1+n2+n3 (8) Integration of eq 8 with the boundary condition (cCl-,0 ) 0) results in



cCl-

0

∫2

dZ ) (cNaClO3,0 - Z)n1+n2+n3

t

n2+n2

‚kdt

(9)

0

1-(n1+n2+n3) (cNaClO3,0 - cCl-)1-(n1+n2+n3) - cNaClO ) 3,0

2n1+n3‚k‚(n1 + n2 + n3 - 1)t (10) or

k)

1-(n1+n2+n3) (cNaClO3.0 - cCl-)1-(n1+n2+n3) - cNaClO 3,0

(11)

2n2+n3(n1 + n2 + n3 - 1)t

Therefore, the global rate coefficient k at a specified temperature can be determined using eq 11 given that n1, n2, and n3 have been obtained, and cCl- at time t is measured. The temperature range, 40-80 °C, important for the FeSO4-based sulfur removal approach, is reasonably narrow, consequently variations of the preexponential factor and the activation energy of reaction 1 with temperature can be ignored (9, 10). Therefore, the relationship between the global rate constant (k) and reaction temperature (T) can be expressed in the empirical Arrhenius form below

(

k ) A exp -

E RT

)

(12)

where A is the preexponential factor, E is the activation energy of reaction 1, and R is the ideal gas constant. The plot of log k versus 1/T can be used to extract values for A and E in Arrhenius form (eq 12), since the slope and interception of such a plot are -E/2.303R and log A, respectively.

Experimental Section

Taking the logarithm of eq 5 yields

log

reorganized as

(7)

The cCl- at any time, t, of the reaction was measured with a Dionex Series 4000i ion chromatograph, which leads to the establishment of the relationship of cCl- vs time. Consequently, the reaction order with respect to NaHSO3, n3, can be derived by curve fitting log dcCl-/dt ∼ log(cNaHSO3,0 - 2cCl-). It is noted that the value of dcCl-/dt at time t was obtained from cCl- - t curve by using a tangential approach (8). Similarly, when cNaClO3,0 and cNaHSO3,0 are much larger than cFeSO4,0, n2 can be obtained by curve fitting log dcCl-/dt ∼ log(cFeSO4.0 - 2cCl-), which is based on the relationship of cCl∼ t attained through experimental data. In addition, when the initial concentrations of the reactants FeSO4 and NaHSO3, cFeSO4,0 and cNaHSO3,0, are much larger than cNaClO3.0, the reaction order with respect to NaClO3, n1, can be inferred by curve fitting the experimental cCl- ∼ t relationship and its corresponding log dcCl-/dt ∼ log(cNaClO3.0 - cCl-) plot. Furthermore, the initial reactant concentrations with the relationship of 2cNaClO3,0 ) cFeSO4,0 ) cNaHSO3,0 were designed to derive global rate coefficient k at a known temperature. Given initial concentrations of three reactants, eq 2 can be

Apparatus. The reactions between NaClO3, FeSO4, and NaHSO3 were conducted in a 1.0 L double-wall glass reactor with three inlets on the top. A stir set through the central inlet of the reactor was used to mix the reactants during the reaction. A thermometer passed through another inlet to measure the temperature in the reactor. The prepared reactant solutions were added through the additional inlet. A DL-501 thermostat was used to control the temperature in the reactor by circulating water around the reactor. Water and solutions of reactants NaClO3, FeSO4, and NaHSO3 were prepared in advance and stored in separate vessels. All the chemicals used for the reactions and analysis of this study were analytical grade. Measurement of Cl- Concentration. The concentrations of Cl- during the reaction at different times were measured for monitoring the extent of reaction. To this end, the reactions in the reactor need to be stopped immediately after a desired period of reaction time. Sodium hydroxide solution (0.1 M) was prepared and used for that purpose, since the reaction between NaClO3, FeSO4, and NaHSO3 stops when the pH of reaction 1 is higher than 7 (11). The pH was monitored using a Fisher Scientific model 310 pH meter. Change of cCl- was monitored for 240 s in each test. Rates for most nonzero and positive order reactions are higher at early times than a later period; accordingly, a shorter sampling interval, 5 s, was chosen for the first 30 s of cCl- monitoring time. This increased the density of the data points in the VOL. 37, NO. 7, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1405

TABLE 1. Cl- Concentrations (cCl-) at 150 s and Global Rate Coefficients Calculated for Reaction 1 under the Different Temperature (cNaClO3,0 ) 14.75 mM, cFeSO4,0 ) 29.5 mM, and cNaHSO3,0 ) 29.5 mM) temperature (°C) cCl- (mM) k (mol-2.6‚dm2.6‚s-1)

40 4.73 17.3

50 7.04 43.3

60 9.02 102.3

70 10.5 230.0

80 11.6 494.0

FIGURE 3. Change of Cl- concentration with time (cNaClO3,0 ) 14.75 mM, cFeSO4,0 ) 29.5 mM, and cNaHSO4,0 ) 0.295 mM, T ) 60 °C).

FIGURE 1. Change of Cl- concentration with time (cNaClO3,0 ) 0.1475 mM, cFeSO4,0 ) 29.5 mM, and cNaHSO3,0 ) 29.5 mM, T ) 60 °C).

FIGURE 4. Determination of reaction order of NaClO3 (cNaClO3.0 ) 0.1475 mM, cFeSO4,0 ) 29.5 mM, and cNaHSO3,0 ) 29.5 mM, T ) 60 °C).

FIGURE 2. Change of Cl- concentration with time (cNaClO3,0 ) 14.75 mM, cFeSO4,0 ) 0.295 mM, and cNaHSO3,0 ) 29.5 mM, T ) 60 °C). initial period of cCl- ∼ t plots, where the cCl- ∼ t curves typically appear steep, and thus improved the accuracy of curve fitted cCl- ∼ t relationship. Ten seconds was used as the time interval for the remaining sampling period. Samples were taken using pipets. The value of cCl- was determined using a Dionex Series 4000i ion chromatograph. The instrument was calibrated with a set of NaCl standard solutions at concentrations of 0, 50, 100, 150, and 200 ppm. Preparation of the Solution for the Determination of Reaction Orders. All the tests designed for the determination of reaction orders were conducted at 60 °C. As previously mentioned, the initial concentration of the reactant for which the reaction order is to be determined should be much smaller than other reactants. Accordingly, 0.4425 mM NaClO3, 88.5 mM FeSO4, and 88.5 mM NaHSO3 solutions were prepared and used for the determination of the reaction order of NaClO3, 44.25 mM NaClO3, 0.885 mM FeSO4, and 88.5 mM 1406

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 7, 2003

NaHSO3 solutions for the determination of reaction order of FeSO4, and 44.25 mM NaClO3, 88.5 mM FeSO4, and 0.885 mM NaHSO3 solutions for the determination of reaction order of NaHSO3. The same volume (150 mL) of each prepared reactant solution (i.e., NaClO3, FeSO4, and NaHSO3) was used in each reaction. Therefore, the real initial concentration of each reactant in each reaction was one-third of that prepared before mixing. Preparation of Solutions for the Determination of Arrhenius Form. Rate constants were calculated at different temperatures in order to determine the Arrhenius equation for reaction 1. To obtain the data needed for these purposes, 44.25 mM NaClO3, 88.5 mM FeSO4, and 88.5 mM NaHSO3 solutions (2cNaClO3,0 ) cFeSO4,0 ) cNaHSO3,0) were prepared and used for reactions at 40 °C, 50 °C, 60 °C, 70 °C, and 80 °C. The concentrations of Cl- generated, cCl-, were measured 180 s after the reactions started. Rate constants at different temperatures were calculated using eq 11.

Results and Discussions cCl- ∼ t Relationships. The results of the tests on the change of cCl- with time under different combinations of initial reactant concentrations are shown in Figures 1-3. The trends of the curves in Figures 1-3 indicate that none of the reaction orders for NaClO3, FeSO4 and NaHSO3 were zero, since none

that the reaction rates of NaClO3, FeSO4, and NaHSO3 are closely correlated with initial reactant concentrations. Derivation of Reaction Orders. The data points in Figures 4-6 were derived numerically from Figures 1-3 using a tangential method (9). The resulting linear plots, Figures 4-6, each have a regression coefficient higher than 0.95. The slopes of the lines are the reaction orders of reactants NaClO3, FeSO4, and NaHSO3 and are approximately 1.1, 1.1, and 1.4, respectively. Therefore, eq 2 can be expressed as follows:

dcCl) k(cNaClO3,0 - cCl-)1.1(cFeSO4,0 dt 2cCl-)1.1(cNaHSO3,0 - 2cCl-)1.4 (13)

FIGURE 5. Determination of reaction order of FeSO4 (cNaClO3,0 ) 14.75 mM, cFeSO4,0 ) 0.295 mM, and cNaHSO3,0 ) 29.5 mM, T ) 60 °C).

Establishment of Arrhenius Form. The average concentrations of Cl- in the reaction system at 150 s under temperature 40 °C, 50 °C, 60 °C, 70 °C, and 80 °C are shown in Table 1. Global rate coefficient k at 40 °C, 50 °C, 60 °C, 70 °C, and 80 °C, calculated with the cCl- values in Table 1 and eq 11, are also listed in Table 1. From these global rate coefficients, it is evident that rate increases with temperature. The global rate coefficient increases by approximately 100% when temperature increases 10 °C in the temperature range tested, which is consistent with van’t Hoff rule (12). Based on the results in Table 1, the relationship between log k and 1/T was determined as demonstrated in Figure 7. A regression coefficient of 0.98 in Figure 7 indicates that a good linear relationship between log k and 1/T exists. Therefore, the Arrhenius form of reaction 1 can be expressed as

k ) 7.2 × 1012‚e-

FIGURE 6. Determination of reaction order of NaHSO3 (cNaClO3.0 ) 14.75 mM, cFeSO4,0 ) 29.5 mM, and cNaHSO3,0 ) 0.295 mM, T ) 60 °C).

8291 T

(mol-2.6‚dm2.6‚s-1)

(14)

In conclusion, the reaction between NaClO3, FeSO4, and NaHSO3, potentially used for sulfur removal from flue gas, was characterized. A kinetic model, including reaction orders and activation energy, has been established by tracking the change of the concentration of the chloride ion with time during the reactions with different initial reactant concentrations. This study is essential in understanding the underlying chemical mechanism of the new sulfur removal technique, which can provide the basis for the design and optimization of the process.

Literature Cited

FIGURE 7. Determinations of activation energy and the preexponential factor in Arrhenius form (cNaClO3,0 ) 14.75 mM, cFeSO4,0 ) 29.5 mM, cNaHSO3,0 ) 29.5 mM, and t ) 150 s). of those plots is linear. All three of the curves, shown in Figures 1-3, rise steeply at the beginning of the reaction and become flat after about 100 s; however, each has a different initial rate (within the first 5 s). From this, one can determine

(1) Xu, Z. Air Pollution Control Technologies, Science Press: Beijing, 1993. (2) Onda, K.; Kato, K.; Kasuga, Y. JSME Intl. J. 1996, 39, 202-210. (3) Fan, M.; Zhuang, Y. Environ. Sci. 1998, 19, 5-8. (4) Fan, M.; Zhuang. Y. Adv. Environ. Sci. 1997, 5, 104-107. (5) Fan, M.; Zhuang. Y. Proc. 1st Asia-Pacific Conference on Sustainable Energy and Environmental Technology, Singapore, World Scientific Publishing: 1996; pp 188-193. (6) Fan, M.; Sung, S.; Brown, R. C.; Wheelock, T. D., Laabs. F. C. J. Environ. Eng. 2002, 128, 483-490. (7) Hendrich, S.; Fan, M.; Sung, S.; Brown, R. C.; Lebepe-Mazur, S. M.; Myers, R.; Osweiler, G. S. Intl. J. Environ. Technol. Manage. 2001, 1, 464-471. (8) Berkowitz, N. An Introduction to Coal Technology; Academic Press: New York, 1979. (9) Mathews, J.; Fink, K. Numerical Methods with MATLAB, 3rd ed.; Prentice-Hall: Upper Saddle River, NJ, 1999. (10) Parravano, G.; Martin, J. J.; Katz, D. L. Elements of Chemical Reaction Engineering, 3rd ed.; Prentice Hall PTR: NJ, 2000. (11) Fan, M. Research on New Desulfurization Technologies, Ph.D. Thesis, Research Center for Eco-Environmental Sciences of the Chinese Academy of Sciences, 1997. (12) Levenspiel, O. The Chemical Reactor Omnibook; Oregon State University Book Stores: Corvallis, 1996.

Received for review May 28, 2002. Revised manuscript received December 18, 2002. Accepted January 13, 2003. ES025828V VOL. 37, NO. 7, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1407